Data Scientist, Product Analytics (Brooklyn, NY (Hybrid) - 1128

In this role, you will be the core analytics partner to our Fulfillment ML, Product, and Operations teams. You will assess delivery ML model performance, analyze end-to-end buyer delivery experiences, and evaluate controlled experiments to measure real-world impact on buyer and seller lifetime value (LTV).

• Advanced Analytics & Experimentation: Partner with ML engineers, operations leads, and product managers to identify model optimization opportunities and evaluate the impact of new features, through randomized A/B testing and offline causal analysis.
• Ad-Hoc Analysis & Exploration: Query, model, and interpret large-scale data in Google BigQuery to answer critical business questions and support strategic roadmapping.
• Metric Frameworks & Data Accessibility: Design, build, and maintain robust SQL data rollups and aggregations to ensure reliable, high-quality data sources for reporting and cross-team consumption.
• Dashboards & Visualization: Create and manage interactive Looker dashboards to track key performance indicators (e.g., on-time delivery, fulfillment friction, resolution rates) and share trends with leadership.
• Cross-Functional Collaboration: Translate complex model behaviors and analytical findings into clear strategic recommendations for non-technical stakeholders.

• Strong background in experimental design, A/B testing, and hypothesis-driven statistical analysis.
• Proven experience building production-ready dashboards and data models in Looker (or similar BI tools like Tableau).
• Advanced SQL proficiency, specifically with cloud data warehouses (e.g., Google BigQuery).

Nice-to-Haves:

• Direct experience analyzing logistics, supply chain, carrier performance, order resolution, or fulfillment incident metrics.
• Experience working in a two-sided marketplace environment
• Experience with causal inference and analytical ML modelling (regression, clustering, random forest, propensity score matching etc.)
• Familiarity with clickstream data (visits, events etc.)
• Familiarity with AI tools (Claude, cursor etc.)